jason.woltjeandClaude Opus 5.5
dbf72a2a3e
Candidate on 96656dd21: broker split refuses invalid-recommendation after the
option checks, raise_decision text asks for one key and names the yes default,
bus README codes table, broker and tools tests. Mutants m1 (old shared line)
and m2 (order swapped) killed; full gate passed 15:42:03-15:47:36Z.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
89 lines
7.0 KiB
Plaintext
89 lines
7.0 KiB
Plaintext
✔ sessionModel: agent vars win, then the system's execution settings (11.585732ms)
|
||
✔ a pi bundle: prompt, policy, tools and manifest, 0600 in a 0700 directory (5.980928ms)
|
||
✔ a claude-code bundle adds the wrapped gate hook and the MCP config (4.216719ms)
|
||
✔ the prompt's built-in tools line is the harness's own names, as the manifest lists them (8.168396ms)
|
||
✔ the prompt names the project clones in the workspace, and none when there are none (5.078342ms)
|
||
✔ the prompt explains the review loop to a role with push_working, and reviews to one with fetch_review (8.106294ms)
|
||
✔ a bundle is written once: an existing file refuses (1.44307ms)
|
||
✔ a path with a single quote can't go into the hook command (1.507379ms)
|
||
✔ allow exits 0, a deny exits 2 with the reason on stderr (116.549776ms)
|
||
✔ a missing or wrong policy, or a bad event, exits 2 (84.27165ms)
|
||
✔ the bundle's wrapped command: a missing gate or node still blocks (1084.581041ms)
|
||
✔ claude: typed tools through MCP, the hook blocks, builtins outside --tools don't exist (713.865671ms)
|
||
✔ claude: the hook alone blocks a path outside the workspace (428.580151ms)
|
||
✔ claude: a second turn resumes the first turn's session (729.251481ms)
|
||
✔ claude adapter: --restricted is always passed (5.262807ms)
|
||
✔ claude adapter: MOSAIC_CLAUDE_BIN wins over PATH and must be an absolute executable (16.969546ms)
|
||
✔ claude adapter: a directory pin exits 2 before it runs; a pin with a space runs (6.510084ms)
|
||
✔ claude: CLAUDE.md files and auto-memory don't reach the model; without --restricted they do (698.797828ms)
|
||
✔ claude: a missing hook or MCP file refuses before claude starts (7.12421ms)
|
||
✔ pi: policy tools and typed tools pass, anything else is blocked (2.858094ms)
|
||
✔ claude: builtins map from pi names, typed tools need the mcp prefix (0.706111ms)
|
||
✔ file tool paths must resolve inside the workspace (1.181646ms)
|
||
✔ pi's own path normalisation can't be used to step out (0.886292ms)
|
||
✔ a symlink inside the workspace that points out is outside (0.734531ms)
|
||
✔ a dangling symlink is refused at any depth, in both harnesses (5.948945ms)
|
||
✔ read is checked under every spelling pi's read would open, in both harnesses (14.485278ms)
|
||
✔ other spellings cover directories, dangling links and pi's cwd (1.623725ms)
|
||
✔ a relative path climbs from the workspace's real path, in both harnesses (5.443554ms)
|
||
✔ claude path fields per tool (0.659905ms)
|
||
✔ glob patterns stay inside the workspace (0.627792ms)
|
||
✔ a path that can't be checked is blocked (0.625538ms)
|
||
✔ fetch_review's have lists the working refs before origin's, and no more than HAVE_MAX (212.517457ms)
|
||
✔ fetch_review's have orders the working refs newest first within each clone, one clone after another (126.920127ms)
|
||
✔ a runner git past its time: SIGTERM reaches its process group, and SIGKILL follows after the grace period (1809.722776ms)
|
||
✔ once a signalled runner git has closed, its group gets SIGKILL, so a process that ignored SIGTERM dies too (304.037362ms)
|
||
✔ initialize, ping and tools/list (41.589177ms)
|
||
✔ tools/call goes through the tool socket; a refusal is an isError result (35.026292ms)
|
||
✔ replies are compact JSON; over the cap a tool error names the size and the cap (45.500909ms)
|
||
✔ unknown tools and methods are JSON-RPC errors and never reach the socket (27.937124ms)
|
||
✔ a missing argument is a usage error (28.272208ms)
|
||
✔ pi: typed tools reach the socket, the gate blocks, agent_end writes the marker (366.1079ms)
|
||
✔ pi: a reply over the cap is a tool error naming the size and the cap; list_tasks asks for open tasks (304.555529ms)
|
||
✔ pi: a write through a dangling symlink is blocked, and nothing appears outside (297.243039ms)
|
||
✔ pi: a read is refused when pi would open another spelling outside (309.384424ms)
|
||
✔ pi: a relative path climbs from the real path of a workspace behind a symlink (304.360525ms)
|
||
✔ pi: a relative path climbs from the real path of a dataRoot behind a symlink (307.921622ms)
|
||
✔ pi: a missing extension refuses before any model call (6.198196ms)
|
||
✔ pi: an extension without its configuration fails pi's start (246.896233ms)
|
||
✔ founderCheck: founder variables, then a needed service without a usable token (1.117151ms)
|
||
✔ turnRequest names the sender, class, reply and decision (0.210064ms)
|
||
✔ turnRequest says a tester's message is a test and never Jason's instruction (0.20187ms)
|
||
✔ turnRequest shows the message's own sent time, a day old as well as new (0.126513ms)
|
||
✔ stateBlock names the launch state with its time and event, and the open tasks with the cap (0.202911ms)
|
||
✔ a message becomes a turn, the answer goes back as a RESULT, SIGTERM releases and exits 0 (239.447176ms)
|
||
✔ launches revoked and then restored: the turn's block reads allowed and gives the restore time (256.991673ms)
|
||
✔ a task closed on the tracker is missing from the next turn's open list (216.781374ms)
|
||
✔ a message read a day after it was sent shows the time it was sent (146.281428ms)
|
||
✔ a failed state read fails the turn: no adapter runs, the sender hears why, the message is read (151.278842ms)
|
||
✔ a SIGTERM before the claim stops the runner with exit 0 and no claim (90.847024ms)
|
||
✔ typed tools carry the runner's capability; launch goes to the host's launch socket (377.545722ms)
|
||
✔ a RESULT gets no automatic reply; failed turns reply with the reason (1284.777166ms)
|
||
✔ SIGTERM during a turn kills the turn's process group and still exits 0 (148.015816ms)
|
||
✔ founder credentials stop before the claim (20) (175.214563ms)
|
||
✔ a refused claim exits 21; an ended run's capability exits 22 (183.651336ms)
|
||
✔ the launch ending under a running session exits 22 (141.019463ms)
|
||
✔ a broker that stays unreachable exits 23 after brokerRetries polls (243.062338ms)
|
||
✔ a broker that is down at the claim exits 23, not 21 (107.177168ms)
|
||
✔ no capability, or a malformed one, on stdin exits 2 (184.852464ms)
|
||
✔ a missing or malformed policy exits 2 before the claim (119.916875ms)
|
||
✔ the PM gets launch, its task verbs and the reads (8.265929ms)
|
||
✔ list_requirements reads the requirement ids, for a task.create holder with a tracker (4.288021ms)
|
||
✔ a coder gets no launch, no resolve_decision, and no task tools without a tracker (1.770926ms)
|
||
✔ a coder with a tracker gets its own task tools and no task_reassign (1.235232ms)
|
||
✔ launch only when the business's launch block names the instance as launcher (1.678299ms)
|
||
✔ an action outside the instance's authority has no tool (1.323397ms)
|
||
✔ the review loop: push_working and request_review for the coder, fetch_review and review_verdict for the reviewer, none for the PM (4.588239ms)
|
||
✔ callTool: one JSON line out, the result back, a refusal rejects (7.473723ms)
|
||
✔ list_tasks takes the broker's filters and pages; get_task takes a task_ref (2.195815ms)
|
||
✔ callTool sends list_tasks done false unless the session names done; other tools as given (2.601207ms)
|
||
✔ replyText: compact JSON up to the cap, an error naming the size and the cap over it (0.730309ms)
|
||
ℹ tests 80
|
||
ℹ suites 0
|
||
ℹ pass 80
|
||
ℹ fail 0
|
||
ℹ cancelled 0
|
||
ℹ skipped 0
|
||
ℹ todo 0
|
||
ℹ duration_ms 11014.388979
|